Can Humanized Mice Support Preclinical Assessment of Anti-Drug Immune Responses?  

Anti-drug antibodies can alter the exposure, activity, efficacy, or safety of biological drug products. Yet conventional animal responses to human biologics can be difficult to interpret because species differences may drive immune recognition.

In a research collaboration between the U.S. Food and Drug Administration and Taconic Biosciences, investigators tested whether NeoThy™ immune humanized mice could support the coordinated cellular and humoral responses involved in anti-drug antibody formation. Published in Frontiers in Immunology, the study found evidence of treatment-associated immune activation, immunoglobulin class switching, antigen-associated proliferation, and binding or neutralizing anti-drug antibodies in selected human donor groups.

The findings provide proof-of-concept evidence that NeoThy mice can mount anti-drug immune responses under the conditions studied. They also define important considerations for determining how such responses may eventually contribute to broader immunogenicity risk assessment.

An anti-drug antibody is the endpoint of a coordinated immune response

ADA formation can involve a connected sequence of biological events. A therapeutic protein is recognized and processed, followed by antigen presentation and activation of responsive T cells. T-cell signals can then support B-cell activation, maturation, and immunoglobulin class switching, ultimately producing antibodies that bind to or potentially neutralize the therapeutic.

An ADA assay measures an important endpoint, but it does not independently describe the adaptive immune process that produced the antibody. Evaluating cellular activation, B-cell maturation, class switching, antigen-associated proliferation, and antibody function together provides a more integrated view.

For drug developers, the relevant question is therefore not simply whether a humanized mouse contains human immune cells. It is whether those cells can act together following exposure to a biological drug product.

NeoThy™ integrates human immune biology with functional assessment

NeoThy mice combine neonatal or pediatric human thymic tissue with human CD34⁺ hematopoietic stem cells. The human thymic environment supports human T-cell development and selection, while the HSCs generate multiple human immune-cell lineages. The mice evaluated in this study were produced on an immunodeficient background expressing human IL-6, which can support aspects of human immune development represented in the model.

Bone marrow–liver–thymus models helped establish the value of incorporating a human thymic environment into immune-humanized mice. However, BLT production relies on fetal tissues, creating ethical, sourcing, or institutional barriers for some organizations. NeoThy provides a non-fetal option, using neonatal or pediatric thymic tissue obtained as discarded tissue from medically necessary procedures and without implantation of human liver tissue.

NeoThy therefore brings together human thymic education, HSC-derived immune reconstitution, human donor biology, and complementary immune measurements as an integrated humanized translational platform.

An FDA–Taconic collaboration tests responses to biological drug products

The proof-of-concept study included six groups of NeoThy mice representing different human thymus and HSC donor combinations. Mice were generated at both FDA and Taconic, and selected starting materials were represented across production sites to help examine the consistency of response. The publication identifies a research collaboration agreement between FDA and Taconic, under which Taconic provided mice and selected study reagents.

Researchers evaluated two biological drug products associated with clinically observed immunogenicity:  

• Infliximab, an anti-TNF monoclonal antibody
• Interferon beta-1b, a therapeutic protein used in multiple sclerosis

Mice received infliximab, IFN-β, both biologics in combination, keyhole limpet hemocyanin as an intended positive-control antigen, or saline. The study integrated longitudinal immune phenotyping with immunoglobulin isotyping, lymphocyte-proliferation assays, ADA assays, and histopathology.

The research collaboration and publication do not represent FDA endorsement or regulatory qualification of NeoThy, and the manuscript is not FDA guidance or policy. They report the findings of a jointly conducted scientific investigation.

Multiple endpoints support a coordinated adaptive response

The mice developed broad human immune reconstitution across blood and lymphoid tissues. The represented populations included T cells, regulatory T cells, B cells, monocytes, myeloid and plasmacytoid dendritic cells, and NK cells.

The study also identified several connected indicators of adaptive immune activity:

• Treatment-associated changes in T-cell maturation and activation
• Activated and class-switched B-cell populations
• Plasmablast development
• Production of human IgM and multiple IgG subclasses
• Antigen-associated lymphocyte proliferation in selected mice
• Binding anti-IFN-β antibodies in selected donor groups
• Neutralizing anti-infliximab antibodies in several donor groups

In responsive animals, ADA formation generally aligned with other measures of adaptive immune activity. For example, mice from one responsive donor group produced anti-IFN-β antibodies alongside evidence of drug-associated proliferation, immunoglobulin class switching, and T-cell activation. By comparison, an ADA-negative donor group remained capable of responding to nonspecific mitogenic stimulation but showed more limited class switching and immune activation.

The significance therefore lies in more than ADA detection. Evidence across cellular, humoral, and functional endpoints supports the interpretation that the antibodies developed within a broader treatment-associated adaptive response.

Human donor biology shaped the response

Responses varied substantially among human donor groups. Donor identity influenced immune-cell composition, lymphocyte activation, immunoglobulin subclass production, antigen-associated proliferation, and whether measurable ADAs developed.

This variability is both biologically informative and experimentally consequential. It indicates that the human immune system represented in the model is an active study variable rather than an interchangeable reagent. It also means that a study based on one donor could miss relevant response diversity.

The findings support including multiple independent donors, adequate replication, and donor-matched treatment and vehicle controls in future studies. However, they do not establish that the observed donor-specific responses correspond to immunogenicity responder profiles in patients.

From anti-drug response assessment to immunogenicity risk assessment

This study establishes that NeoThy mice can support preclinical investigation of anti-drug immune responses. It does not yet establish that the platform can predict clinical ADA incidence, determine a product’s overall immunogenicity risk, rank candidate molecules, or quantitatively reproduce patient responses.

Immunogenicity risk assessment is broader. It integrates the likelihood and potential consequences of an immune response with product attributes, therapeutic mechanism, dosing, patient factors, nonclinical evidence, and clinical data.

With further fit-for-purpose validation, NeoThy findings could provide one biologically integrated evidence stream within that broader assessment. Comparative studies across larger donor panels and biologics with different clinical immunogenicity profiles will be needed to determine where the platform can most reliably inform specific drug-development decisions.